A dozen figures shuffle through a false darkness, most wearing night-vision goggles that no one needs. In Afghanistan's dry autumn air, even a slender scimitar moon casts sharp shadows. But the younger soldiers feel twitchy. And this village, called Malajat, has just become a proving ground for America's new approach to war.

Malajat lies south of Kandahar and dead center in the conflict. The road courses with dust a foot deep in places, so fine it ripples away from boot heels. "We call it moon dust," murmurs a Special Forces major named Fernando Lujan. Unlike the other soldiers, he wears a long black beard that frames a lean, dark face and black eyes. He speaks Dari. He could duck down any corridor and blend into the village if he wanted. Instead he follows in silence as the younger soldiers – mostly military police, fresh from home – make a spectacle of night patrol, fidgeting with their goggles and signaling each other with absurd bird noises. Up, down. Off, on. Coo-ee! Coo-ee!

Lujan exhales through his beard. The 33-year-old Texan leads a separate, elite team of just three men who regularly insert themselves into Afghanistan's worst battlegrounds, by whatever means, often alone. They are dispatched by General David Petraeus himself, commander of the war in Afghanistan, to size up whether his counterinsurgency tactics are working. Those tactics have triggered an identity crisis within the military and have divided some of the thinkers and strategists who plan America's wars. Counterinsurgency is widely credited with turning around the Iraq campaign, but traditionalists nonetheless consider the approach dangerous and delusional. Unlike conventional tactics, it calls for field diplomacy as much as firepower. If American soldiers can win over the Afghan people, the theory goes, the Afghans will stand up for themselves against the Taliban. Now Lujan's team will watch that theory in action and return to Kabul to stand before Petraeus and report success or failure.

So this platoon in Malajat is a test case, in a sense, and so far Lujan doesn't like what he sees.

The MP squad's leader, a young sergeant, wants to introduce himself to the most important person in Malajat: the new village mullah. Having set up a security perimeter at both ends of the block, the squad leader knocks on the door of the central mosque.

A teenage boy opens the door and gapes at the small crowd of American soldiers on the stoop, each bristling with weapons, radio antennae, rifle magazines, and grenades, with eyes illuminated by the green glow of nocturnal vision.

"HELLO!" the MP bellows.

His translator offers, "Salaam alaykum."

The teenager nods back, and the MP sergeant launches into a memorized set of questions. Is the new mullah home? When will he be in? What's his name? Where did he –

Before long, the translator and the boy speak between themselves. The teenager speaks Pashto, and it turns out the translator, unbeknownst to the MP, is a Dari-speaking Hazara. He struggles with Pashto. Eventually, he turns to the MP. "He says this is not a good time. It is time for evening prayers. Maybe if we could come back in an hour?"

The MP freezes for a moment, searching his mind for a contingency plan. Finding none, he moves back to the top of his list: Is the new mullah home? When will he be in? Where is –

A man in his 30s joins the young man at the door. The new mullah, according to the translator. Maybe. It's complicated. The MP frowns. "What's your name?"

"Al-Abdul."

The MP frowns harder. "We had another name."

The young mullah stares. "It's time for prayer," he says. "Please."

The sergeant hesitates, but he returns to his script. "Welcome to Malajat!" he says, and fires new questions. He is the conversational equivalent of an assault rifle. Between bursts, the stars seem to spin overhead, so long and awkward are the silences. Lujan drops back from the scene and leans close to his civilian teammate, a former SEAL and black-ops specialist who goes by the name Sam. "So this is going well," Lujan whispers, and Sam offers a wincing smile.

At the mosque doorway, the MP continues what has become Operation Small Talk. The young mullah, clearly torn between his religious responsibilities and the intimidation of the soldiers, finally breaks away in the middle of a question. He storms off to issue the evening call to prayer. As his voice rings out over Malajat, the squad seems unnerved by the sound and falls into disarray. Some soldiers lag behind, while the others move down the street. Coo-eee!

Lujan shakes his head. His work lies ahead of him with this crew.

Until just a few years ago, America's military focused on tactics any veteran from World War II could recognize: Identify an enemy, apply all available weapons to kill that enemy, and then withdraw, leaving the nation-building to politicians.

In the 21st century, though, America's enemies have made an abrupt shift. On a battlefield crowded with civilians, the military struggles to even identify the enemy. Its high-powered, long-range weapons can't distinguish a radical madrassa student from his moderate classmates. So counterinsurgency's aim is not to wipe out the enemy, but rather to drive a wedge between the population and the insurgents. This creates space for a legitimate government to flourish and isolates the extremists for a more surgical kill. And that, in the end, creates fewer enemies.

"This isn't just the thinking man's war," says John Nagl, a retired lieutenant colonel and a leading counterinsurgency proponent. "This is the graduate level of war." But opposing officers and soldiers say counterinsurgency, or COIN, is a mockery – a sabotage – of the very nature of warfare. "We've essentially turned much of the United States Army into a mix between constabulary forces designed to police unhappy Muslims, and a nation-building corps." That's retired Colonel Doug Macgregor, who says his disdain for counterinsurgency was one thing that prevented him from becoming General Doug Macgregor. "I wonder how they would perform if we suddenly had to fight against someone with real capability? I don't think we would fare very well."

It's an old dilemma. As long as governments have ruled, insurgents have rebelled. In the first century A.D., in North Africa, a Roman Army deserter named Tacfarinas led a rebellion against the Roman Empire. Rather than stand nose-to-nose against the superior Roman legion, Tacfarinas and his band of Berbers blended into the population. He knew Rome's commanders could sort the sands of the Sahara itself before they could sift through the tribal complexities of the region. To kill the rebels they would have to kill everyone.

Much later, British archaeologist-turned-soldier T.E. Lawrence led uprisings in Arabia, Syria, and Jordan on behalf of Britain. He wrote that, in his experience, any insurgency that could enlist just 2 percent of the population as combatants could be very difficult to defeat. "War upon rebellion was messy and slow," he wrote, "like eating soup with a knife."

Here's the trick of counterinsurgency: The hardened 2 percent depend on at least passive sympathy from the other 98 percent. Mao Tse-tung, the leader of the Communist revolution in China, studied Lawrence's tactics and wrote of the local population's relationship with insurgent fighters: "The former may be likened to water, the latter to the fish who inhabit it."

Tacfarinas held off the legion until a Roman commander named Blaesus made a critical discovery: Instead of slashing and burning his way across the countryside, as usual, he offered the war-weary tribesmen a deal. If they turned in their weapons, they would receive a full pardon from the Empire, and Blaesus's troops would protect local villages from any revenge by Tacfarinas in his retreat. And it worked: The population ended its tacit support for the rebels. With deft encouragement from Blaesus, the sands had sorted themselves.

The question, now, is whether that's possible today.

David Kilcullen, architect of modern warfare, sits ensnared in Kabul traffic. It's a chaotic scene that feels more like a drive-through bazaar than a road. A pack of Afghan children tap at the car window, and one boy waves a canister that billows smoke. The boy swears it's incense and will bring good luck – a pipe bomb of happiness. The children gawk at Kilcullen's red hair and ruddy complexion, not noticing that his left hand stays hidden beside his thigh, where it holds a Glock pistol.

"This isn't good," Kilcullen says, his Australian accent perpetually rising. He explains: A band of beggar children is like a neon arrow, pointing the Taliban's spotters toward a Western target. Traffic is bad. Children are bad. The two together can be explosive.

Kilcullen should know. He serves, officially, as a lieutenant colonel in the Australian Army inactive reserves. Unofficially, he is one of the most influential people in the theory and practice of counterinsurgency, and he has helped recast the modern American military, from high doctrine down to the battlefield.

When we reach our destination – an officers school for counterinsurgency, outside Kabul – Kilcullen springs from the car to meet the school's new commander. Colonel Chad Clark is a sharp-eyed veteran of elite counterterrorism units, and when Kilcullen mentions the traffic and children, Clark does something unexpected: He grins.

It turns out Kilcullen had read the street scene wrong. He would have been spot-on a year ago. But insurgency is so fluid, so unreliable, that the opposite is true now. "Anywhere you see a traffic jam, things are all right," Clark says. "Children are a good sign. People pull their kids inside when they sense something's about to happen."

That's good news, Kilcullen says, brightening. Things are looking up in Kabul.

Kilcullen has come to Afghanistan from his base in Washington, D.C., to talk strategy with General Petraeus. The two are good friends, and the general – "Dave" – looks to the 43-year-old Kilcullen for advice. Both have a lot at stake in counterinsurgency: If Petraeus manages to successfully end two of the country's longest wars, the hardest part of any potential presidential campaign would be behind him.

Even though a third of the country is still under the influence, if not outright control, of the Taliban, Kilcullen is ebullient. He has a genius for breaking down any challenge into a series of clear, simplified steps. The first step toward victory here – or at least a dignified exit – is undoing the damage wrought by conventional warfare over nearly a decade. And he is not above trash talking. He caused a stir in 2008 when he called the invasion of Iraq "fucking stupid." It might not have mattered, except he worked at the time as a special adviser to Condoleezza Rice.

Now, in Kabul, he's feeling jaunty again. "We cleaned up their mess in Iraq," he says of the military's old guard, "and now we'll do it here."For a whole generation of Americans, the very word counterinsurgency had the tang of vulgarity, when it was spoken at all. Our last attempt at it, in Vietnam, left a lingering bad taste. And if traditionalists had their way, counterinsurgency would have died a flaming death in Saigon.

So the latest thinking on counterinsurgency took root in foreign soil, far from the Pentagon. In 1993, Kilcullen, then a young captain and anthropologist in the Australian Army, moved to a West Java city to study the Indonesian language. One day, he noticed a military museum there, and inside he found a display about a decades-old war between the Indonesian government and a Muslim insurgency called Darul Islam. Somehow Kilcullen – a student of military history – had never heard of it, and he wondered why. His generation knew counterinsurgency only as an utterly failed idea. But here, under glass, he had found evidence that it could succeed on a large scale: The Indonesian government had handily defeated the rebel movement. He threw himself into researching the conflict and eventually focused his doctoral thesis on counterinsurgency as applied against Darul Islam.

He made this discovery at a time when almost no one else knew, or cared, about counterinsurgency. So by day, Kilcullen served as an Australian officer in Indonesia, and in the evenings, he visited the nation's aging rebel leaders who'd been put down by the government. They resisted at first, but over time, and over tea, he pressed them for details on tactics, strategies, motivations, weaknesses.

A few years later, Kilcullen led a company of peacekeepers protecting East Timor residents from retaliatory violence as they made a break for independence. To casual observers this Christian-separatist insurgency bore little resemblance to the Darul Islam rebellion 50 years earlier. But Kilcullen realized the two movements, religion aside, shared critical tactics and motives. And if a counterinsurgent force could isolate those elements, it could break a rebellion.

He finished his disseration in 2001, just as the world started to care – deeply – about counterinsurgency.

Over the next few years, Kilcullen watched the world's finest armies spin in the sands of Afghanistan and Iraq. One night in 2006, he sat down with a pen, a notebook, and a bottle of Laphroaig scotch. Toward the end of World War I, one of Kilcullen's heroes, T.E. Lawrence, had written Twenty-seven Articles, a note to other British officers about how to lead an insurgency. So that night Kilcullen answered with Twenty-eight Articles, about how to end one.

Insurgency and counterinsurgency pivot on the same point. Lawrence included it in Article 15: "Do not try to do too much with your own hands. Better the Arabs do it tolerably than that you do it perfectly. It is their war, and you are to help them, not to win it for them."

Kilcullen wrote in his Article 13, "Conduct village and neighborhood surveys to identify needs in the community – then follow through to meet them, build common interests and mobilize popular support. This is your true main effort: Everything else is secondary."

The message is plain, in both cases: Don't die on someone else's soil if you can help it. Win over the locals and train them to wage their own war. If winning them over requires digging wells in the meantime, so be it. Lightbulb wiring at the local mosque? Fine. Baby formula for the orphanage? Sure.

There weren't many new ideas in Kilcullen's Articles. Other thinkers had made similar points before. A half century ago, for instance, in the midst of widespread French failure in the Algerian War, French officer David Galula used COIN tactics to smash the insurgent forces in his area. At about the same time, British envoy Sir Robert Thompson used similar means to put down a rebellion in colonial Malaya.

Those men were brilliant field commanders. But Kilcullen's gift is fermenting the ideas of past intellectuals and distilling them to their most potent, palatable essences. He thinks in terms of shapes – circles, arrows, pillars, and arches – that lend themselves to diagrams. I once saw him, over lunch, sketch the entire Afghan conflict on a scrap of paper and then solve it with one tiny arrow. "That's where we're working to break the cycle," he said. Such simplicity can be intoxicating.

Kilcullen sent out his own 'Twenty-eight Articles' by email to a few colleagues and tacticians scattered around the world, who then passed it along to their superiors and subordinates, who forwarded it further, until the Australian's note became a sort of idealogical pinup girl, stuffed in rucksacks, tacked to corkboards, scrutinized for every blemish and beauty mark. Lawrence's original Articles showed up in 1917 in a publication called the Arab Bulletin and took generations to seep into popular thinking about war. But Kilcullen's email, within months, landed on the desks of some important military thinkers.

One of them was David Petraeus, who at the time was plotting a coup of his own within the U.S. military. After years of watching America's wars in Afghanistan and Iraq slowly slide toward failure, a group of young strategists – military, civilian operators, diplomats – felt their superiors were dangerously out of touch. The old guard had grown up waging a supersize, purely theoretical Cold War. They didn't know how to stop an insurgency. So that's exactly what the younger subordinates staged: an insurgency.

"It was a bottom-up rebellion, by junior officers with combat time in the field," Kilcullen says. "That's why I had to write 'Twenty-eight Articles' over a whiskey in the middle of the night – it was an underground document. Petraeus was our leader and top cover."

At the time, the Army kept Petraeus relatively marginalized, along with his ideas about counterinsurgency. He was in charge of Fort Leavenworth in Kansas, where he oversaw several schools and training programs. But he also – critically – oversaw the assembly of a new handbook for how America wages war. The field manual is much more than a guide or collection of lessons learned; it is the rudder that steers the ship. It is doctrine. Petraeus assembled a cohort of COIN students to write it, including Kilcullen. The Australian's seemingly breezy, simple approach appealed to Petraeus; it brought order to the fur ball of ethical, political, and tactical questions Americans faced in Afghanistan. "As an insurgency ends," Kilcullen wrote, "a defection is better than a surrender, a surrender better than a capture, and a capture better than a kill." And just like that, with the publication in late 2006 of FM 3-24: Counterinsurgency, the COIN advocates overthrew the traditional U.S. military. Soon Petraeus took charge of the war in Iraq and brought Kilcullen on as a top adviser. That's how, less than a year after sending his middle-of-the-night email, Kilcullen helped shape and direct the Iraq troop surge of 2007.

The speed of Kilcullen's ascent and the adoption of his ideas says much about the flexibility of modern war, about his own clarity of thinking, and about America's desperation for a solution in Iraq and Afghanistan at the time. Along the way, though, a fissure formed in the American military bedrock. For a while, Kilcullen said, "Donald Rumsfeld actually banned us from using the words insurgency or counterinsurgency – mainly because for the guys who invaded Iraq, that would have been an admission of failure."

Now some soldiers and scholars ask: America's finest and deadliest are handing out lightbulbs? Baby formula? For crying out loud.Deep in the windowless archives of the U.S. Army Center of Military History in Washington, D.C., a lone figure hunches over a musty binder of Vietnam-era communiqués. Stacks of binders surround him, giving the impression of a man taking cover inside a hastily built fort. "You won't find Dave Kilcullen in here," he says, grinning.

Colonel Gian Gentile is, you could say, counter-counterinsurgency. He did two tours in Iraq, rising to cavalry squadron commander, teaches at West Point, and has emerged as a leading critic of COIN's role in U.S. strategy. All the new fence mending and hand shaking, he says, will atrophy the military, whose reason for existence – whose sole reason – is killing, without apology.

It's not that Gentile likes big, bloody wars. His political views are "far left," a product of growing up in California, he says, laughing. He's a large man with a tight haircut, but he wears a congenial, lopsided smile and swears more than he means to. He wears a carved wooden bracelet, a symbol of his wife's Buddhist beliefs. He's a hodgepodge. But he loves his country, he says, and too much counterinsurgency has "crippled" America's forces. "The American Army today has become so consumed with counterinsurgency tactics that COIN tactics and operations have now eclipsed strategy."

Strategy, in military terms, is a leader's vision. Tactics are the tools used to accomplish it. Clearing pirates from the Horn of Africa is an example of strategy; sending SEALs in inflatable Zodiac speedboats is a tactic. Gentile says America's leaders in Afghanistan have confused the two, and that gets him riled.

"When I hear a four-star general in Afghanistan, like Petraeus or [former Afghanistan war commander General Stanley] McChrystal, talking about separating the insurgents from the population," he says, "that's the language of tactics and operations. That'd be like Eisenhower in the summer of '44, giving a speech on strategy, but on how infantry squads were clearing 88 [calibers] from the hedgerows."

What option do the current generals have, though? Politicians made the decision to go to war, so the strategy is already set in motion. Good tactics are the only way out.

"How do you end a war that's lost because of failed strategy and policy?" Gentile says. "Do you just keep pushing and hope that you'll get better at tactics and operations, and that'll save you?" You can't persuade people to defend themselves, he says. "It's all such bullshit, man. This whole fucking fetish with behavior and influence, and that men with guns can somehow sway people's minds and change their behaviors."

That's the philosophical abyss that separates COIN advocates and opponents. Counterinsurgency advocates believe in war as a sharp instrument. Their opponents view war as a sledgehammer, a blunt and heavy object designed for killing.

Those opponents appeal to history. Counterinsurgencies lose, they say. Look at Algeria, site of the French debacle. Or Vietnam.

But, respond the COIN advocates, Galula defeated insurgents in his part of Algeria because he understood counterinsurgency. And what about Northern Ireland? Or Malaya? Oman? Or countless smaller counterinsurgencies in Colombia, El Salvador, Peru, western Sahel, the Philippines, and elsewhere? Yes, counterinsurgencies lose, but 60 percent of them are successful, COIN advocates claim. And consider the closest analogy we have for Afghanistan: Iraq. Didn't counterinsurgency work there? Petraeus dangled the carrot of dollars and swung the stick of increased ground operations, and violence plummeted. Now the place is finally climbing to its feet.

That claim touches a nerve when put to Gentile. "We don't know how Iraq is going to turn out," he snaps.

With that, the colonel returns to his binders. They hold reams of cable communiqués from Vietnam war commander General William Westmoreland and his successor, General Creighton Abrams. Westmoreland embodied the traditional approach: a hard-charging, hammer-swinging leader who used search-and-destroy tactics that focused on the enemy. Abrams favored counterinsurgent methods, focused on winning the hearts and minds of the population. History remembers Westmoreland poorly for his role in Vietnam, and Abrams as the general who would have rescued victory if he hadn't run out of time. Gentile feels otherwise. "People think we were losing in Vietnam, and oh, a better general with better tactics came in and saved the day," he says, waving his arms for emphasis. "Nonsense."

That's what led Gentile to dig through antique war correspondence from two dead generals. "There was more continuity than change in Vietnam after Abrams arrived," he says – people have it backward. And in a way he's right: Westmoreland once declared that the jungles of Vietnam were "no place for either tank or mechanized infantry units." And Abrams – well, the Army named a tank after the guy.

Abrams, Gentile feels, showed up just in time to snatch the scraps of glory. And Gentile hates that.

In 2006, before the surge, Gentile's cavalry squadron took a beating in Iraq, but it handed out more casualties than it suffered, by far. And still the observers back home and abroad said the Americans were losing in Iraq. Then a new general – Petraeus – showed up with an Australian adviser – Kilcullen – just before the war there turned around.

And now they want to do the same in Afghanistan.

Fernando Lujan, the special forces operative, peers from the rear window of a car as Kandahar Airfield slides past. It's a jarring scene.

In the past decade Kandahar has split into two worlds: inside the wire, and outside the wire. Inside the wire, there are taco nights, and big-screen televisions, and volleyball games, and wide roads with sensible speed limits, and Internet cafes – mile after mile, all surrounded by a ring of walls topped with razor wire and well-armed soldiers.

Outside the wire, there is only danger and death.

So as Lujan watches us pass the outer perimeter, he gives a half smile and says, "And there goes the wire." We're in a civilian car, wearing Afghan clothes and head scarves wrapped to obscure. Lujan slouches with a particular Afghan boredom as the car enters the desert. "Now we're in Afghanistan. The real Afghanistan." Sam, the civilian operator on Lujan's team, wears the grim but pleased look of a prison escapee. In the front seat, the Afghan driver smiles. We're on our way to Malajat.

Commanders throughout history have battled the wire, the bureaucratic isolation, struggling to interpret reports as they burble up through the ranks. Petraeus knows that. So in the face of an old problem he has reached for an old tool, developed before the technologies of drones, or satellites, or even radios, all the way back to Napoleon Bonaparte. The French general craved information and employed a tactic now called the "directed telescope," allowing him to peer through the fog and smoke of war. The telescope was a small team of brilliant young officers who operated without regard to rank, roamed across the battlefield, and then reported their findings directly to Napoleon. Good news or bad, he heard it in firsthand detail. And now, in Afghanistan, Petraeus has a directed telescope of his own: Lujan and his team.

As a human telescope – part of the Afghan Counterinsurgency Advisory and Assistance Team, or A-CAAT – Lujan sees the fullness of the war, from low-down gut shots on the battlefield to the strategies and pressures of command. On one day he'll teach basic English to Afghan fighters from far-flung provinces. The next he'll advise top strategists on the course of the entire war. Lujan arguably has a deeper understanding of the war – and counterinsurgency – than anyone else in Afghanistan.

Our car approaches a checkpoint manned by paramilitary guards with weapons. "We try not to stop," Lujan says. Better to avoid questions whenever possible. At the checkpoint's bottleneck, the Afghan driver gives a wide smile and a familiar, dismissive wave, as though to say, "No need for formalities among old friends." He keeps rolling through the checkpoint, betting correctly that the guards, wherever their loyalties lie, will not bother to shoot.

Lujan describes his job as a calling, and it is hard to imagine anyone more suited to it. He was born to a father he never met – an illegal Mexican immigrant who bailed out before his birth – and a mother incapable of caring for him alone.

He grew up among other relatives and their consorts, with no prospects for anything better. So he delivered a shock, one day as a teenager, when he announced that he had applied to, and been accepted by, the United States Military Academy. West Point, he told them.

He worked hard as a cadet. Then as a young soldier he became a Ranger, and then a member of the elite Special Forces, which led to missions around the world, including in Iraq. Then he returned to school – to Harvard, actually – for another degree and a better grasp of international affairs. He took a position teaching at West Point but soon grew disillusioned with academia; war had changed, but West Point hadn't. The academy taught America's young officers obedience, yes, but never prepared them for the flexible, fluid world of modern war. How could they learn to adapt, when even the number of inches between clothes hangers in their closets was dictated to them? Lujan felt called to do more – for his country, yes, but also for his students. So he volunteered for service in Afghanistan.

He still writes to his students, feeding them practical knowledge. "Insurgents have figured out how to make caveman-simple, pressure-plate or command-wire devices that all our technology is worthless against," he said in a long, heartfelt note a couple of months into his deployment. "We're seeing a reverse evolution in tactics, and they're deadly effective."

The note offers blistering, field-view insight into counterinsurgency, which military professors and philosophers could study as closely as any set of articles. "Every tiny piece of terrain out here exacts a terrible price," he wrote. "You just hope that commanders will learn quickly to pay only once. Don't clear what you can't hold. The Taliban will reclaim it before you finish walking or driving back to your base and lay IEDs along the way you came."

When Lujan left West Point, he published a paper that called for a revolution at the academy. It should stop manufacturing rule-memorizing bureaucrats, he said, and start fostering innovation. He lost friends and colleagues after that. His students heard his character questioned. So he ended his note from Afghanistan with an explanation of his decision. "Sometime in every single one of your military lives, you will face a moment where you see something that you deeply disagree with and believe needs to be addressed." When that day comes, he wrote, he hopes they'll break the rules. And accept the sacrifice that follows.

Now Lujan watches the Afghan landscape scroll past his car window, at ease wearing what lesser soldiers call "hajiflage." An assault rifle rattles at his knees and a sack of hand grenades sits at his feet. He has little interest in the debate among thinkers back home, on theories and traditions. He's a practician now. And he has more pressing matters at hand.The village of Malajat is famous, in certain circles, because in all the years the Soviets occupied Afghanistan, they never took this small clutch of farmland and mud homes.

It's an infantryman's nightmare. It sits on land irrigated by the Arghandab River, so soldiers face thick plots of corn stalks and whole forests of marijuana. But the worst, by far, is the vineyards; without much wood for trellises, farmers here grow their grapevines on rows of thick mud walls that can reach eight feet tall. The narrow paths through the vineyards are natural sites for booby traps, so every movement becomes an exhausting affair, as soldiers heave themselves and their gear up and over wall after wall after wall.

Lujan and Sam arrive in the middle of the night and are shown to a tent bustling with maps, satellite equipment, radios, and technicians lit by computer screens. They receive a briefing from Captain Ethan Olberding: Coalition and Afghan forces recently cleared Malajat and are now entrenching themselves with the local population.

Olberding is, by all appearances, an action figure. A small flashlight beams from a strap around his perfectly bald head, and when he talks, the features of his face – his eyeballs, his jaw – jut forward. He uses a voice so commanding it sounds like an enormous child has pulled a string in his back to wind him up and make him speak. He is not a commander, in short, who seems inclined to win over anybody's heart or mind.

The next night Lujan and Sam join the discombobulated MP unit and a couple of Afghan paramilitary policemen for the night patrol that leads them through the moon dust to the village mosque. After the bungled exchange with the young mullah, the group pushes forward with worsening results: They stop and search almost everyone they see, from children to an old man bent over his cane. During one exchange with a shopkeeper, the unit's leader doesn't bother to flip up his night-vision goggles, baffling the man. While they speak, gunfire pops about two kilometers away, which in this village is equivalent to a dog barking in the distance. Even so, one of the squad members comes trotting in: "Sir! We've got gunfire! Gunfire, sir!" The leader strides off – away from the gunfire – while the shopkeeper hangs in mid sentence. As the unit scrambles to pull itself together, Lujan approaches the shopkeeper with his hand over his heart, and after a few clarifying words, they chuckle and part ways with a handshake.

When the platoon files back into its compound, one of the Afghan police partners – ignored by the MPs throughout the patrol – pulls at Lujan's sleeve. Then he releases a stream of indignant Dari: The Americans were rude toward the shopkeeper, he says. Who searches old men and small children? Not to mention the disaster with the mullah. "He says it alienates the local people," Lujan translates. "He's embarrassed."

Sam gives a short, curt nod. "Tell him we want him to say all that again, in a moment. Exactly that way."

The MPs toss off their gear and gather to smoke and laugh and discuss the night's adventure. From the darkness, Lujan and Sam approach, then they introduce the Afghan cop, who gives his polite but frustrated report. "Every one of these mistakes could have been avoided if you had just consulted your Afghan partners first," Lujan says. "The partnership has to be real. The sooner we hand off to the Afghan authorities, the sooner we go home." The MPs straighten a bit at the word home. They start asking questions. Lujan has gotten through.

Not every squad reacts that way. During an earlier patrol, an infantry unit had performed fine, but it showed no curiosity about the territory it patrolled. So afterward, Sam, who has decades of experience operating in foreign cultures, stood before them and offered practical tips for gathering intelligence – a brief master class from an expert in the subject. "Think of yourselves as Lewis and Clark," he said. "When Thomas Jefferson sent them to explore the Louisiana Purchase, they weren't sure what to expect. So they took down every detail, filling in the picture for the president."

As he talked, a chubby medic in sunglasses stood near Sam and lit a Swisher Sweets cigar. He then blew long, luxurious clouds of smoke into Sam's face. Sam ignored him. When the brief tutorial was finished, the troops wandered away and someone offered a sarcastic, "Verrry interesting!"

After Sam's talk, I ask the cigar-smoking medic, whom Lujan later dubbed "Mr. Sweets," how he feels about counterinsurgency tactics. "Stupid. They ought to use that money in American schools," he says. He pokes his cigar at several spots around the compass. "What is the point, here?"

His anger illustrates a contradiction in practical counterinsurgency. When an average 19-year-old heads to war and leaves his hometown, mom, girlfriend, dog, and everything else, he has to ratchet up his mind, somehow, to find the ability to kill people. One way is to dehumanize the population. So when it comes time to pull the trigger, that little twinge, that hesitation he might feel back home, doesn't get in the way. It's an effective technique. Not so long ago the U.S. government used it as a matter of outright doctrine, with wartime cartoons and caricatures of Japs and Jerrys. And in a total war, in which all a soldier needs to do is shoot, it might work.

But how does the young soldier unwind that mind-set long enough to give some Afghan farmer a second glance? Because now he can't just accept the Afghan's enmity as part of his base nature; he's got to figure out the man's grievance. Maybe he's growing poppies on his farm only because the Taliban threatened to kill his family. Is he a Taliban sympathizer? Or a supporter? Maybe a potential ally?

After the night patrol, Lujan and Sam visit Captain Olberding – the company's commander – in his operational hub. They offer a candid report: The MPs need help. They need closer leadership. Olberding drops his smooth head into his hands and kneads the skin over his eyebrows. "All right," he says, then looks up again at the bank of computer screens before him. He doesn't have time for fuzzy counterinsurgency advice. "Thanks. I've got some things I need to do here."

Over and out.The next day the MPs gather for another patrol, and – to everyone's surprise – Olberding emerges from his tent and approaches with several Afghan policemen.

"So, we've got some work to do," he says. His voice booms as though from a hidden speaker. He plans to join the patrol, he says, as do the Afghan police, whom he introduces by name. And – strange – the captain carries several prayer rugs rolled under his arm.

"What we're doing here matters, guys," he says. He explains the historic significance of Malajat, and how Petraeus himself asks for news from the village each week. And in this hard-won little town, in one of the most critical provinces in all of Afghanistan, the local mullah is the most important figure.

The group files into the street, and Olberding leads them back toward the mosque – in daylight, this time – waving and nodding. "Salaam alaykum!" he booms to everyone who passes. An old man peeks from his doorway as Olberding stops at the mosque and asks to meet the mullah. The young man comes to the door again, looking hesitant. Olberding apologizes for the confusion of the night before and offers the prayer rugs as a small gift. The man's posture opens a bit, and as time passes he warms to the two squad leaders, American and Afghan. Then a particularly Afghan thing happens: From a doorway across the street the old man emerges and approaches the knot of soldiers. He introduces himself: He, not his young helper, is the village mullah. And thank you, he says, for the prayer rugs.

Lujan watches as the conversation grows, and finally Olberding poses a question Lujan had suggested earlier: Some of the Afghan partners have no place to pray. Would the mullah mind if they came here, to the mosque?

Not at all, the old man says. They are welcome.

In the background, Lujan beams. A night ago the squad had blundered through Malajat, manhandling the very villagers who could help expose the Taliban. And now, after a gentle nudge from Lujan and Sam, they've gained entry to the town's central mosque. A search-and-destroy mission could never have opened that door.

Back at the compound, Lujan reflects on this small victory. He sacrificed a comfortable career as a West Point professor to come here, to a place where he spends most nights sleeping outside. But war has changed, possibly forever, and he couldn't keep teaching young officers outdated lessons. "Consider the way we operate," he says. "Even when we construct our small operating bases in the middle of nowhere, our units are quick to separate themselves from Afghan counterparts, to build up infrastructure – Internet, satellite TV, their own living quarters, et cetera – to surround themselves with reminders of home and Western society. We construct a tiny 'bubble' wherever we go."

That's the ideological ambush that waits in the space between traditional training and David Kilcullen's beautifully clear view of the Afghan war, with circles and arrows. It's so clean, so simple, that the average 19-year-old can grasp it. Then he arrives in Afghanistan, where beggar children may or may not pose a threat and the village mullah has two names. Suddenly there are degrees of Taliban-ness, and the messiness of motivations, and hesitations at every turn. And his training fails him.

Counterinsurgency is more than the thinking man's war, as John Nagl said. It is war that requires a man to keep two discrete minds: He must befriend with his left hand and kill with his right.